I sent a check-point trickle yesterday at 19:54, did an update at 2:00 today (UTC+1) and got no credit, so my statistics graph has taken a nose-dive. I was jusr getting it levelled out.

It looks OK on boincStats, but note that boincStats can be a day behind.

You\'ll never have a flat \'average\', since once it reaches a natural level it will \'sawtooth\'. Credits are generated at roughly a daily interval so it\'ll be some time before it\'s obvious whether theres a problem or not.

---

I'm a volunteer and my views are my own.
News and Announcements and FAQ

Last I heard, the credits as displayed at this site are calculated once a day, I think sometime between 6 and 12 UTC.

I\'ve been sending trickles at about the same time every 48 hours, and the tops of the sawteeth were beginning to show a consistent value. Then suddenly comes this big dip where no credit has been granted. So I was worried. The graph does now show an upturn, but not to where it should be. Perhaps today\'s trickle will put it back on course.

If you\'d been roughly synchronised with the credit generation you may have drifted to just before it rather than being after it. In optics this effect is called moire interference (two periodic processes drifting in and out of sync at regular intervals).

If you were to \'update\' more frequently, say, several times per day, then you\'d see the sawtooth pattern appear more clearly. The pattern is a jump upwards when the credit calculation is done, followed by an exponential decay.

---

I'm a volunteer and my views are my own.
News and Announcements and FAQ

Oh yes, I do update several times a day. The 48 hours I mentioned is when a \"check-point\" trickle occurs, i.e. one of those trickles every 10,802 steps which generates credit. I have arranged the resource share so that this occurs in the evening (about 19:00 UTC every 2 days) when I am on line and can watch it happening.
Then at the next update after midnight UTC the Stats graph gets a horizontal extension, which changes the next day to an up or a down, alternating to produce the saw-tooth. In this particular instance I unusually got 2 consecutive downs, which has thrown the graph awry, and indicated that the credit was not generated within the usual time.
I wonder why Climate have chosen such a short average-calculation time, whereas others use a longer span, e.g. I complete an Enstein unit every 3 days, yet the graph is dead straight. Climate is the only one that saw-tooths.

I wonder why Climate have chosen such a short average-calculation time, ...

This is because a year ago, a lot of multi-project people who were only giving cpdn a small percentage of their total time, were complaining that they were at a disadvantage, compared to people who crunched cpdn exclusively.
So the formula was altered, and my rac, (and that of others who only crunch cpdn), dropped from over a thousand to a couple of hundred.

(And now it\'s zero, as I\'ve been crunching \'special\' models for 3/4 of a year, and these projects have their one set of credits.)